The increase in lumbar spine BMD in response to Raloxifene (RLX), a selective estrogen receptor modulator, is smaller in magnitude compared to the response to treatment with estradiol (E2). The reasons for this observation are unclear. Estrogen has a potent effect on the production of proinflammatory cytokines which support osteoclastogenic and bone resorption. Therefore the different response to RLX may relate, at least in part, to a difference in the ability of RLX to modulate the production of proinflammatory cytokines which are abundant in the red marrow of the vertebrae. The aim of this study was to determine the effect of RLX and E2 both in vitro and ex vivo on the production of the pro-resorptive cytokine interleukin-1beta (IL-1beta) and its antagonist, interleukin-1 receptor antagonist (IL-1ra). We obtained samples of peripheral blood from (a) 10 untreated postmenopausal women with osteopenia (ages 53 to 72 years, mean 61 years), (b) 15 postmenopausal women (ages 52 to 72 years, mean 63 years) at baseline and after 6 months of RLX therapy (60 mg/day) and (c) 10 postmenopausal women (ages 60 to 75 years, mean 64 years) at baseline and 6 months after a single E2 implant (25 mg). Cultures of whole blood from the untreated women were incubated with RLX or 17beta-E2 at 1 pM, 100 pM, 10 nM and 1 microM concentrations. LPS-stimulated whole blood cultures from the raloxifene- and estradiol-treated women were prepared at baseline and at 6 months. IL-1beta and IL-1ra were measured by ELISA in the conditioned media. In vitro there was a significant dose-dependent decrease in IL-1beta and IL-1ra in response to 17beta-E2 (both P<0.0001) which was not apparent in response to RLX (both P>0.05). In ex vivo cultures from women receiving 6 months treatment with E2 implants, there was a significant decrease in IL-1beta (-36+/-8%, P=0.01) but no significant change in IL-1ra (+29+/-20%, P=0.3). There was no significant change in either IL-1beta or IL-1ra after 6 months RLX therapy (+20+/-14% and +12+/-10%, both P>0.05). We conclude that treatment with RLX, unlike estradiol does not modulate the production of the proinflammatory cytokines IL-1beta and IL-1ra using in vitro or ex vivo whole blood culture methods. This may account, at least in part for the reduced efficacy of RLX therapy compared to estrogen which has been observed in vivo on bone mineral density, bone turnover and reduction in fracture risk.